Table 3 Average performance obtained from the external validation of SVM models on independent datasets developed using different types of features.

From: Prediction of Immunomodulatory potential of an RNA sequence for designing non-toxic siRNAs and RNA-based vaccine adjuvants

Features (Composition) Thres Average Performance of Models
Types Vector size Sen Spec Acc MCC
MNC 4 0 78.42 ± 5.26 79.71 ± 3.38 79.02 ± 2.44 0.58 ± 0.04
DNC 16 0 90.50 ± 3.60 89.42 ± 2.03 90.00 ± 2.43 0.80 ± 0.05
TNC 64 0 92.58 ± 2.27 88.65 ± 2.02 90.80 ± 0.82 0.82 ± 0.02
TetNC 256 0 92.83 ± 2.52 90.58 ± 2.47 91.79 ± 1.56 0.84 ± 0.03
PNC 1024 0 95.00 ± 2.39 91.73 ± 2.09 93.48 ± 1.49 0.87 ± 0.03
5′ bin 68 0 89.83 ± 2.88 88.85 ± 2.91 89.38 ± 1.83 0.79 ± 0.04
3′ bin 68 0 91.25 ± 3.00 85.96 ± 3.15 88.79 ± 1.57 0.78 ± 0.03
5′-3′ bin 136 0 93.92 ± 3.47 90.67 ± 3.82 92.41 ± 3.09 0.84 ± 0.06
hybrid 200 0 93.50 ± 3.23 91.44 ± 3.19 92.54 ± 2.34 0.85 ± 0.05